Extracellular space preservation with immersion fixation of brain sections

Here’s a nice article from 2021: “Permeabilization-free en bloc immunohistochemistry for correlative microscopy” by Fulton and Briggman. A few thoughts:

1. The protocol helps to overcome an impediment to correlative en bloc EM and IHC by enabling the ultrastructural preservation of brain tissue and antibody penetration into relatively thick tissue sections, without the use of permeabilizing agents. This could help to facilitate the use of pre-embedding IHC in ultrastructural analysis techniques such as 3D EM.

They replicate the finding that permeabilization dramatically decreases ultrastructural quality, so should be avoided if possible:

https://elifesciences.org/articles/63392, Figure 1

They argue that a key way they were able to accomplish antibody penetration without permeabilization was via the preservation of the extracellular space (ECS).

2. The brains sections they used are still quite thin relative to those that are practical in neuropathology, at 300 um – 1 mm. In neuropathology, human brains are somewhat frequently sectioned at 5 mm, but even that is challenging and requires expert skill.

One could try to use a device such as a compresstome to help with the sectioning process. Here’s a video showing how the compresstome works on mouse brains. But it seems difficult to scale this to human brain sizes.

(If one were to achieve such thin sections in a high-fidelity way, you could theoretically cryopreserve them with vitrification or near-vitrification procedures and therefore avoid fixation altogether. Although, avoiding fixatives would also make room temperature preservation not currently possible.)

3. Another possible mechanism for why their protocol worked, that the authors did not discuss as far as I could tell, is that tissue decomposition during the immersion fixation process — which is slower than perfusion fixation — may itself cause membrane permeabilization. With a long enough time period of decomposition, cell membrane breakdown is an inevitable event, so the question is really whether the immersion fixation was slow enough to allow it to occur. My guess is that it was a contributing factor.

This may also help to explain why some epitopes are more accessible (eg Homer) than others (eg PSD-95). If a protein is a part of stronger gel-like networks, this gel-like network will likely break down slower during the decomposition process, and therefore be more difficult for antibodies to access without permeabilization.

4. Do we even need immersion fixation for ECS preservation? They cite Cragg 1980 as an example of a study that achieved ECS preservation using perfusion. It’s still not entirely clear to me why perfusion doesn’t usually achieve ECS preservation, but it seems like it probably depends on the osmotic concentration of the perfusate. Cragg 1980 is 30+ years old now; it would be ideal if it could be replicated and the phenomenon understood better.

Prenatal epigenetic age acceleration in Down syndrome

That’s a result of Xu et al 2022, “Accelerated epigenetic aging in newborns with Down syndrome”.

This study furthers our understanding of a syndrome of accelerated aging. The authors show a significant acceleration of an epigenetic aging marker in the blood of people with Down syndrome. Furthermore, they show that this effect is present at birth and is significantly stronger in newborns who have Down syndrome plus GATA1 mutations. This association with GATA1 mutations is intriguing as GATA1 mutations are associated with transient abnormal myelopoiesis. One thing that this study does not do is investigate the mechanism by which this age acceleration occurs.

One hypothesis based on this finding is that it might help explain why people with Down syndrome have an increased susceptibility to Alzheimer’s disease. Lore has long been that this is due to the triplication of amyloid precursor protein, however, this study suggests that age acceleration may also play at least a part in the increased susceptibility of people with Down syndrome to aging-associated cognitive impairment and Alzheimer-type neuropathology.

from https://onlinelibrary.wiley.com/doi/10.1111/acel.13652

Correlating immunohistochemistry with serial block-face electron microscopy of neurons

In Talapka et al 2022, “Application of the mirror technique for block-face scanning electron microscopy”, the authors use a modified “mirror” technique to combine immunohistochemistry for labeling of dendrites and ultrastructural analysis in 3D-EM of osmicated sections. This relies on the finding that the surface of a tissue block can still be imaged using confocal microscopy. The authors show that the cell body of a somatostatin immunopositive neuron and one of the emerging dendrites can be clearly visualized and reconstructed after the use of their technique. It is likely that the dendritic arbor of a large number of neurons can be analyzed using this technique. The technique combines the advantages of a high-resolution approach and of a labeling method for specific cellular markers. The morphological preservation of the structures seen on the surfaces of tissue sections such as blood vessels will in part determine the quality of the images. Here is one of the figures from their paper:

image from https://link.springer.com/article/10.1007/s00429-022-02506-w

Integrating synchrotron microtomography with electron microscopy in the study of mammalian brain tissue

Bosch et al 2022, “Functional and multiscale 3D structural investigation of brain tissue through correlative in vivo physiology, synchrotron microtomography and volume electron microscopy”, is an interesting study that brings X-ray microscopy to bear on the problem of correlating structure and function.

The authors studied hippocampal CA1 and olfactory bulb circuits via multiple imaging modalities, including 2-photon calcium imaging, X-ray microscopy, and serial block-face electron microscopy. In all cases, the imaging modalities had different strengths in identifying different circuit elements, and the authors were able to correlate structure and function in interesting ways. The interplay between structural, functional, and molecular-level data will be increasingly critical in systems neuroscience, and this study highlights some important points.

The authors should be commended on showing that X-ray microscopy can be used without causing significant damage on fixed and osmium/uranium/lead en-bloc EM embedded tissue, which is an important advance. The authors also showed that X-ray microscopy can be used at high resolution on thick mammalian brain tissues; this is important because X-ray microscopy has the potential to provide structural details at the level of individual dendrites, which is possible with volume electron microscopy but less easily scalable. Finally, the authors point out that staining protein and lipid distributions defines the ultrastructure of the tissue; this is an important point that is often missed.

Figure 4 from Bosch et al; https://www.nature.com/articles/s41467-022-30199-6

Question #21: How far would a typical molecule diffuse in a millisecond?

What is diffusion?

Diffusion is a type of passive transport that involves the net movement of molecules or ions from an area of higher concentration to an area of lower concentration down a concentration gradient. The concentration gradient is the difference in concentration between two points.

In biology, diffusion plays an important role in many biological events such as molecular transport, cell signaling, and neurotransmitter movement across a synaptic cleft.

How far would a typical molecule diffuse in a millisecond? A second? An hour?

Diffusion is a description of how molecules will randomly move around in a liquid. Their movement will be limited if they hit a barrier or randomly collide with another molecule and react, which is not described by diffusion.

The distance a molecule will diffuse in a certain amount of time depends on the size of the molecule, the viscosity of the fluid, and the temperature.

This can be explained by the Stokes-Einstein relation: D = kT/(6πηa), where:

– D = the diffusion constant

– k = the Boltzmann constant

– T = the temperature

– η = the viscosity coefficient of the fluid

– a = the radius of the diffusing molecule

The constant value is 6 assuming that the radius of the diffusing molecule is greater than the radius of the solvent.

Assuming that we are talking about diffusion at 25° C and in water, then there is a nice calculator on physiologyweb.com that lists diffusion coefficients for different ions and molecules:


If we are talking about the diffusion of a small molecule neurotransmitter such as glutamate, it has a MW of 147, which is close to glucose’s MW of 180. So we can use glucose’s diffusion coefficient as a rough guide for the diffusion of some types of small molecule neurotransmitters.

This calculator suggests that glucose will diffuse 1000 nm in a millisecond, 31,000 nm (31 μm) in a second, or 1,900,000 nm (1.9 mm) in an hour.

Molecular diffusion rates are helpful when building intuition about what structural information is necessary to be able to infer in brain preservation. Because, in the way that I think about it, molecular events that occur more slowly than rapid long-term memory recall can be instantiated (which, conservatively, can occur in ~500-1000 ms) cannot be uniquely necessary for the structural information describing it.

Inspired by CalTech’s Question #21 for cognitive scientists: “What is diffusion? How far would a typical molecule diffuse in a millisecond? A second? An hour? How does the diffusion equation differ from the cable equation?”

Question #20: What does the cable equation mean for neurons?

A simplified explanation of capacitance in neuronal membranes is that higher capacitance will tend to will cause a flow of ions towards the membrane on the cytoplasmic side due to the difference in charge across the membrane, called a displacement current; https://en.wikipedia.org/wiki/Cable_theory#/media/File:NeuronCapacitanceRev.jpg

– Cable theory can be derived in part from Ohm’s law, the fundamental theory of electricity that models the current flowing between two points as equal to the voltage distance between the two points divided by the material’s resistance, or in other words, the classic equation V = IR.

– The greater the cross-sectional area of the neurite’s cytosol (the interior part of it, containing biomolecules, electrolytes, and other ions), the easier an ion can flow through it, so the neurite’s longitudinal resistance, r_l, will be lower.

– If the cell membrane is more resistant to ion flow into or out of the cell (due to high membrane resistance, r_m), then charge will tend to accumulate inside the cell, and it will have a higher ionic current flowing down longer distances in the neurite. This is often represented by a paremeter called the length current, λ, equal to the square root of r_m divided by r_l.

– If a cell membrane has a lower membrane capacitance (c_m, which is usually a fairly constant value), then the relative ion flow down the neurite will be greater, due to a lower displacement current. How quickly the membrane voltage changes in response to a current injected at at given point can be predicted by the time constant, τ, equal to the product of c_m times r_m.

– An electrotonic potential results from a local change in ion conductance, e.g. after a synaptic event, that does not propagate. It becomes exponentially smaller as it spreads. This is opposed to an action potential, which reaches the voltage threshold by which it does propagate down the neurite (due to the opening of voltage-gated ion channels), and then spreads like a wave.

– Dendritic trees can perform non-linear integration of signals that can be predicted on the basis of cable theory. The existence of subthreshold membrane potential fluctuations in dendrites, which based on my understanding should dominate neuronal signaling, can allow variations in synaptic weight distributions and input timing to encode a substantial amount of information within a single neuron.

Inspired by CalTech’s Question #20 for cognitive scientists: “Derive the cable equation (for a uniform cylinder, with optimal boundary conditions). What does it mean for neurons?”

Question #19: Ion channel biophysics

What are the biophysics of voltage-gated sodium channels? 

Sodium channels are a major component of excitable membranes. They are an intrinsic component of excitable tissue that allows them to generate and propagate action potentials. These electrical signals are essential for proper neuronal communication.

The channel looks like a barrel, with 4-fold symmetry, and a diameter of about 10 nm. The channel has an activation gate, through which sodium ions can flow through. If the activation gate is closed, no ions can pass through, but if it is open, ions can pass through the pore. The channel is closed at rest, wherein the membrane voltage potential is polarized. When a sufficient voltage depolarization across the membrane occurs, the membrane will draw the gates open, allowing sodium ions to flow through and leading to further depolarization. When enough sodium has passed, the further voltage change causes the inactivation gate to close, thus stopping the flow of sodium ions and leading to repolarization.

Sodium channels are selective for sodium ions because the inner filter of the pore is highly negatively charged; the Na+ ion has a positive charge and will bind well to the inner filter. K+ ions, while also positively charged, cannot pass through because of a size restriction. The gate is not large enough for them to fit through. For ions to pass, they need to be smaller than the diameter of the filter; for ions with a larger diameter to pass, the filter would need to enlarge; however, the size of the filter cannot increase because the pore has a fixed size. These are the unique properties of the sodium channel that allow it to selectively conduct sodium.

Sodium channels are good targets for many drugs and toxins. For example, tetrodotoxin specifically binds to voltage-gated sodium channels and can stop sodium channels from opening, thereby blocking all neural signaling. 

What are the biophysics of transmitter-gated channels? 

Transmitter-gated channels are opened by transmitters. They are then generally ion-selective. To open the channel, the transmitter needs to bind to the receptor. The transmitter binding causes an allosteric change that allows another part of the channel to open, known as the ion channel gate. When open, the ion channel gate allows specific ions to pass through.

A special example is the NMDA receptor. Under normal circumstances, the NMDA receptor is blocked by Mg2+ and Zn2+ ions. When the post-synaptic neuron is depolarized, however, Mg2+ and Zn2+ ions are repelled. In this case, the receptor can be activated by glutamate. When activated, the NMDA receptor allows positive ions to pass through (K+, Na+, and Ca2+ ions), which can help sustain depolarization and lead to intracellular signaling events such as long-term potentiation.

NMDA receptors are often called “coincidence detectors” because these two events must occur together for the channel to open. First, the NMDA receptor must be activated by the post-synaptic being depolarized, and second, glutamate must be released.

Another example is the nicotinic acetylcholine receptor. When acetylcholine binds to the receptor, the channel opens. This allows sodium and potassium ions to pass through, which leads to depolarization and therefore a neural signal.

Most types of ion channel activity in the brain need regulation. Regulation can occur post-translationally through the addition of a phosphate group to one or more amino acids. The addition of a phosphate group to a particular location of the AMPA receptor, for example, has been shown to increase the probability of AMPA channel opening. The Ca2+/calmodulin kinase II pathway is able to phosphorylate the GluA1 AMPA receptor subunit at Ser831, causing an increase in AMPA channel conductance.

In addition to the post-translational regulation of channel activity, many channels are regulated by endogenous compounds in the brain. Serotonin is a monoamine neurotransmitter that regulates various types of sodium channels and potassium channels. Dopamine is also a monoamine neurotransmitter, and it can be found in extrasynaptic regions. Dopamine has been shown to increase potassium channel activity by activating dopamine D1 receptors in axons.  

Together, the biophysics of ion channels allow for neural signaling by allowing for the passage of ions into and out of the cell. This allows for changes in membrane potential and intracellular signaling. 

Inspired by CalTech’s Question #19 for cognitive scientists: “Describe the main biophysical characteristics of ionic channels. How does its biophysical properties contribute to its physiological function? What is thought to be the basis for the channels ion selectivity?”

Microwave irradiation as a brain banking tool

I love random reading old papers. They’re like treasure chests into secret knowledge that others might overlook. Here’s one: Guidotti et al 1974, “Focussed microwave radiation: a technique to minimize post mortem changes of cyclic nucleotides, dopa and choline and to preserve brain morphology”.

As a summary, they found that microwave irradiation in rat and mouse brains for 2 seconds led to the total inactivation of the enzymes involved in the regulation of numerous brain metabolites, such as cyclic AMP, cyclic GMP, choline acetyltransferase, and phosphodiesterase. As a result, it allowed for the accurate dissection of different brain nuclei for measurement of the concentrations of a variety of different metabolites in a variety of brain areas.

Rapid inactivation of enzyme activity; https://doi.org/10.1016/0028-3908(74)90061-6

The microwave irradiation method described by Guidotti et al. 1974 was relatively crude, as the authors only described microdissection of brain nuclei. It is unclear whether delicate cellular features such as synapses might also be preserved.

Interestingly, microwave irradiation produced extensive denaturation of proteins throughout the brain. Because since enzymes control metabolism, the inactivation of these enzymes can minimize changes in metabolism.

Microwave radiation could theoretically be combined with other brain preservation methods, such as fixation or cryopreservation, to minimize enzyme-driven autolysis during the procedure. A potential limitation of this method is that microwave heating is limited by the thermal conductivity of the tissue.

What can super-resolution light microscopy tell us about biomolecular brain preservation?

A lot. A nice example of the power of aldehyde fixatives to preserve fine molecular detail is Helm et al 2021.

Dendritic spines, which are often considered the functional units of neuronal circuits, strongly vary in size and shape. This study used electron microscopy, super-resolution microscopy, and quantitative proteomics to characterize > 47,000 spines at > 100 synaptic targets, helping to quantify variation in biomolecular composition across spines. Their study is amazing in part because of their technical advances, which allow for the beautiful visualization of biomolecules across neuronal membranes.

People often say that connectomics is not enough for brain information preservation because each dendrite has its own spread of ion channels. This distribution of ion channels will tell you whether a dendritic spike will occur, which is incredibly important to figure out synapse function.

If the local dendritic tree goes over a certain threshold of depolarization, then the local ion channels will open up and amplify what would have come in with the synapses alone. This also could synergize with clusters of synapses.

Theoretically, each neuron could have a unique spread of ion channels along dendrites, which could potentially make synaptic connectivity data alone insufficient, even if you have or can accurately infer synapse molecular information.

It’s hard to find examples of real evidence in the literature for or against how important these effects are. But the nonlinear effect of clusters of synapses is something that we potentially can’t account for with electron microscopy data alone.  

This is a reasonable/principled objection to the idea of brain information preservation via connectivity. Personally, I find it quite plausible. A way to address this objection is to say that super-resolution microscopy techniques like those used by Helm et al 2021 could be applied to decoding memories from fixed brain tissue via measuring biomolecules, without necessarily assuming that synapses alone will be sufficient.

Biophysically realistic neuron modeling in layer four of visual cortex

I recently checked out the interesting article by Arkhipov et al 2018 and wanted to discuss it here. They use a well-grounded computational model of L4 (which is the input layer) of mouse visual cortex that is capable of replicating a number of experimental observations. Their model combines biophysically detailed neuron models, synaptic dynamics, and experimentally constrained connectivity. Here is their summary figure describing the biophysical model as well as the leaky integrate and fire (LIF) portion:


This model is effectively supposed to summarize much of what is known in the entire field. It is in line with the Markram approach to brain modeling: we know that from electrophysiology data how these subcircuits of V1 L4 cells work and now let’s use this prior to knowledge to build a Hodgkin-Huxley based model of it. They assessed their model performance by reproducing a number of experimental findings, such as:

1. They reproduced the statistical features of V1 neuronal responses, such as their log-normal distribution.

2. They systematically investigated how neurons in the model respond to a variety of visual stimuli, both the type of stimuli mice might see in the real world (movies) and ones they would not (gratings).

3. As expected from previous literature, they showed that connectivity rules strongly impact neuronal responses. For example, adding recurrency to the network not only amplifies and synchronizes firing rates, but also biases the neuronal tuning properties.

While not directly studying memory storage, the Arkhipov et al study is relevant to the brain preservation problem in a number of ways.

Primarily, it shows that compartmentalized circuits, such as L4 of V1 can be simulated in accurate ways using contemporary biophysical models.

It also highlights the likely importance of connectivity rules in memory storage. Connectivity patterns form the backbone of any computational model, and Arkhipov et al demonstrated how connectivity rules can shape (and constrain) network activity. Therefore, this is another data point that we must consider connectivity patterns to be crucial factors in memory storage.

Their study used compartment models (“compartmental representation of somato-dendritic morphologies (~100–200 compartments per cell) and 10 active conductances at the soma that enabled spiking and spike adaptation”). Their results are a data point that at the single-neuron level, this amount of information may be sufficient for simulations able to reproduce in vivo functional properties, therefore suggesting a potentially reduced need for fine-scale detail preservation, although this is of course still subject to considerable uncertainty.

They also compared the performance of their model to a much simplified version, where biophysical neuron models were replaced by point-neuron models with either instantaneous or time-dependent synaptic kinetics. They found that the biophysically realistic model had quantitatively better accuracy compared to the experimental model, although even with extreme simplification, the model still performed fairly well. This suggests that a level of model detail even above the compartment models may be sufficient.

While they got their connectivity patterns in a random fashion, one might imagine instead getting connectivity data from an electron microscopy-based connectomics data set. It would be interesting to see if, in a realistic biophysical model and a realistic connectomics data set, they could still reproduce a similar set of functional observations.

Theoretically, using these kinds of models, if one were to go beyond just L4 of V1 to a whole brain, it is interesting to think what kind of functional properties might emerge. However, I personally think we should be judicious about building such models.

(Thanks to Ken Hayworth for a discussion about this paper.)